[llvm] [LV] Enable considering higher VFs when data extend ops are present i… (PR #137593)

[Sushant Gokhale via llvm-commits]

sushgokh wrote:

> It sounds like you are looking for shouldMaximizeVectorBandwidth?

Maybe. Looking at the code for AArch64, I think it just considers if vector is FixedWidth. But other than that, there are no other conditions.
This patch adds the condition for this decision. Maybe this can be refined now or later.

> @huntergr-arm was working on enabling it for SVE but it can lead to issues with cost-modelling and perforance, which I believe were being worked through.

I tested the patch on Neoverse-v2 with SPEC17 and there are no regressions. Are there any specific cost-modelling/performance  issue you know of?

https://github.com/llvm/llvm-project/pull/137593